LUNAR IRON—EXAMINATION OF IRON ABUNDANCE ESTIMATES VIA THE M3, CLEMENTINE, AND LUNAR PROSPECTOR DATA SETS FOR ACCURACY AND OPTIMIZATION

Several near-infrared algorithms have been developed to quantify lunar surface iron abundances with reasonably good accuracy. Some of these algorithms use visible wavelengths while others leverage visible and near-infrared wavelengths with reasonable success. However, despite a general agreement among the algorithms, differences remain between the algorithms across various lunar surface terrains as well as with iron abundances derived from the Lunar Prospector Gamma Ray Spectrometer (GRS). As a result, interpretation of iron abundances from any single method, near-infrared or gamma-ray, may imply somewhat different geologic materials, lithologies, and/or igneous or geomorphic surfaces processes. This leaves the knowledgeable analyst in a somewhat difficult position to decide which iron estimate is most reliable. Recent hyperspectral near-infrared data collected by Chandrayaan-1’s Moon Mineralogy Mapper (M³) provide an additional data set to examine lunar surface iron abundance estimates and reconcile with other near-infrared and gamma-ray data sets. An initial application of algorithms is carried out with existing published parameters. However, the near-infrared algorithms are predominantly intended and optimized for deriving iron concentrations from Clementine visible and near-infrared datasets. Here we examine application of three iron estimate algorithms on M³ data and compare the results to iron abundances derived from Clementine near-infrared and Lunar Prospector gamma-ray data sets. We further begin to derive optimization methods that achieve self-consistency among all the datasets.